TEST #1780 RESULTS
11/12/2025 - 5:31 AM
ACCELERATOR
MODEL
352
tokens/s
generation
61
ms
time to first token
24782
tokens/s
prompt
5229
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD EPYC 7R13 Processor (znver3)
RAM
61.9GB
OS
Linux
Kernel Release
6.14.0-1016-aws
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #16~24.04.1-Ubuntu SMP Tue Oct 14 02:15:09 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
35676
tokens/s
364
tokens/s
31
ms
pp4096+tg256
16550
tokens/s
312
tokens/s
251
ms
pp2048+tg256
25705
tokens/s
336
tokens/s
83
ms
pp2048+tg768
25707
tokens/s
336
tokens/s
83
ms
pp1024+tg1024
35583
tokens/s
353
tokens/s
32
ms
pp1280+tg3072
33430
tokens/s
332
tokens/s
41
ms
pp384+tg1152
34734
tokens/s
370
tokens/s
14
ms
pp64+tg1024
11844
tokens/s
386
tokens/s
8
ms
pp16+tg1536
3814
tokens/s
377
tokens/s
7
ms